1 / 22

HP NonStop Server (Himalaya) Business Continuity Customer Experiences

HP NonStop Server (Himalaya) Business Continuity Customer Experiences. Moderated by Harry Scott, Carr Scott Software Presenting: Barclays Bank – Phil Nye Principal Consultant Brown & Co. – Robert Cline Vice President, a division of JP Morgan Chase & Co.

raine
Download Presentation

HP NonStop Server (Himalaya) Business Continuity Customer Experiences

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. HP NonStop Server (Himalaya) Business Continuity Customer Experiences Moderated by Harry Scott, Carr Scott Software Presenting: Barclays Bank – Phil NyePrincipal Consultant Brown & Co. – Robert ClineVice President, a division of JP Morgan Chase & Co. BBS Norway – John GreenPrincipal Consultant Session – ES-02-U

  2. Presentations and Challenges • Barclaycard UK -- Phil Nye • Barclaycard's business continuity environment with emphasis on their use of TMF with ACI's Base24 • Brown & Co. USA -- Robert Cline • Brown & Co's disaster recovery model with emphasis on Indestructible Scalable Computing and ZLE • BBS Norway -- John Green • BBS Norway's computing environment with emphasis on complete environment replication

  3. Phil Nye via speakerphone from the UK

  4. Barclaycard – BASE24 • Barclaycard – A large BASE24 Installation • Nearly 7 million transactions peak day. • Over 230 Transactions per Second • Around 280,000 POS Terminals • Switch to Contingency in single-digit minutes • Barclaycard – A long History of using BASE24 and TMF/RDF • Late 1980’s: Added TMF support to BASE24 POS3.0 • Throughout 1990’s: BASE24/XP – written for TMF • 2000-2001: Standard BASE24 POS5.3 / AutoTMF

  5. The Problem ? How to provide a contingency solution for BASE24 POS 5.3 and ATM 6.0 ? What did we do ? Used 2 * S72014 systems and stress-tested BASE24 POS 5.3 (Beta Release) with: • Golden Gate’s Extractor / Replicator • Compaq's NonStop AutoTMF

  6. So what did we conclude ? • AutoTMF does add TMF support to our high-volume BASE24 configuration. We were able to stress-test a single BASE24-POS 5.3 Logical Network to 250 tps. • We saw significant improvements in disc I/O: • PTLF: Entry-sequenced: DP2 writes up to 56K (=40 TPS) • Able to keep e.g. PTLF/ILF Alternate Keys in Cache • Minimises the impact of Block Splits to e.g. SAF files • POS Settlement cutover is some 50% faster • You can also add TMF support to your Batch programs

  7. Production Experiences: • Started to migrate from BASE24/XP to BASE24-POS and AutoTMF in mid-2001. Fully live by 4th Qtr 2001. • NO problems experienced in production with AutoTMF. • Because of the way AutoTMF determines transaction boundaries, we run with more TMF transactions than Business Transactions. • Forced us to implement RDF IMP/x v 1.2 (to eliminate RDF Receiver Transaction Status Table problems). NO problems experienced with this version of RDF – and we ship Format 2 PTLF files. • At peak we were running with some 650 TMF txns/sec and generating some 1.4Gbytes /sec to the TMF Auditrail discs (1 MAT and 2 Auxiliary Trails).

  8. BASE24 and AutoTMF specifics: • We don’t AUDIT the following files: • NEGs (can’t FUP RENAME an audited file). • IDF (problem with the way BASE24 handles fileinfo call – AutoTMF can now hide all details of an Audited file from an application) We use AutoTMF to AUDIT the following files: • P/TLFs and Alt.Keys • ILFs and Alt Keys. • PTDF/PRDF • SAFs • Database (CPF, etc.) • A VERY simple AutoTMF configuration. ALL GLOBAL parameter settings and WEAK AutoTMF Isolation mode. • We run multiple BASE24 Logical Networks • We don’t RDF-protect all of these (our choice) • We don’t RDF certain files (e.g. ILFs) – reduces Expand requirements • Those files not RDF’ed are configured to use Auxiliary AuditTrails

  9. Presentations and Challenges • Barclaycard UK -- Phil Nye • Barclaycard's business continuity environment with emphasis on their use of TMF with ACI's Base24 • Brown & Co. USA -- Robert Cline • Brown & Co's disaster recovery model with emphasis on Indestructible Scalable Computing and ZLE • BBS Norway -- John Green • BBS Norway's computing environment with emphasis on complete environment replication

  10. Brown & Co. • Front and back office Brokerage system • SIS system (ADP/SIS) • Computer environment, application and procedures upgrade • Latest and greatest Himalaya technology • Business continuity and business intelligence initiative • Sub 10 minute business continuity • Leverage business continuity investment for business intelligence • Adding TMF and SQL to an Enscribe, non-TMF, legacy 3rd party application package (ahead of the vendor)

  11. An Integrated Approach to Replication

  12. \PRIMARY \REMOTE \TEST 12 Processor 12 Processor 6 Processor S74000 S74000 S74000 4 Full APP and 1 Full APP and 1 Full APP and 2 Full REP 3 Full REP 2 Full REP Environments Environments Environments Project Lifecycle (PLC) and Replication from a Platform View REMOTE PRIMARY REPLICATION Site Site REPLICATION PRODUCTION ENVIRONMENTS PERMANENT WALL TEMPORARY WALL CODE MIGRATION DEVELOPMENT AND TEST ENVIRONMENTS \DEVELOPMENT 6 Processor S74000 CODE MIGRATION 6 Full APP and 3 Full REP Environments REPLICATION REPLICATION

  13. Brown & Co.'s Uses RDF for... • Less than 7 minute full fail-over with 100% data integrity to ensure Business Continuity • Foundation for 'no downtime, no excuses' as part of ISC and CPA initiatives • Duplicate database used for data mining as part of ZLE implementation • Duplicate database is source for 'off-host' replication

  14. Presentations and Challenges • Barclaycard UK -- Phil Nye • Barclaycard's business continuity environment with emphasis on their use of TMF with ACI's Base24 • Brown & Co. USA -- Robert Cline • Brown & Co's disaster recovery model with emphasis on Indestructible Scalable Computing and ZLE • BBS Norway -- John Green • BBS Norway's computing environment with emphasis on complete environment replication

  15. Bankenes Betalings Sentral (BBS) • eGiro system • Messageway (BCE Emergis) • MCI (Messageway Clearing Interface) & “waiting register” database • S74008 & S74004 • Kabal project • AutoTMF, AutoSync, RDF & TACL • First joint user of AutoTMF and AutoSync in Europe • First to implement DR for Messageway with RDF

  16. What was replicated • Database Replication • Messageway – Audited and previously non-audited files • MCI & Waiting Register - All audited • Exception is one sequence number allocation file • All transient, intermediate files • Application Infrastructure files

  17. How did we replicate the application ? • NonStop AutoTMF / RDF • AutoTMF Enabled BBS to protect non-audited elements of the Messageway database • TAL, COBOL85, NMCOBOL, NMC, pTAL and NMCPLUS code • BCE Emergis supplied utility handles system name changes

  18. How did we replicate the application ? • NonStop AutoSYNC • Source, Object, TACL macro files • TACL macros handle CPU number and system name changes • Transient files, written with Bulk I/O • Did not use AutoTMF for performance reasons • AutoSync was a key success factor for the project • Saves money in the longer term: No manual procedures necessary and likelihood of errors eliminated • TACL macros handle CPU and system name changes

  19. Automated Replication Procedures • Handling “Parked Data” • Data repository of unstructured files • Index for each record with pointer to file & RBA • Data repository written by single process using Bulk I/O • Read by multiple readers as permanent extended data segment • Could non be audited, so AutoSync used to transfer input files to backup system • Input files re-processed on independent Parked Data subsystem • Communicating TACL macros running on primary and backup systems handle housekeeping functions

  20. AutoSync Usage • Closely monitor performance • Consider multiple batch ID’s • Data compression should generally be disabled • Does not replicate purges of empty filesets (may need dummy files)

  21. Summary • Barclaycard UK - Phil Nye • TMF (and AutoTMF) make applications run faster and provide greatly increased data integrity (including recovery from human failure) • Brown & Co. USA - Robert Cline • RDF provides the fastest, most reliable database replication for disaster recovery, available and is the foundation for ISC / CPA • BBS Norway - John Green • Replicating your database isn't enough, AutoSYNC provides complete application protection from disasters, ensuring that the application environment is in place, in addition to what RDF provides for your databases • Copies of this presentation can be found at:www.CarrScott.com

  22. Want More Information? • Compaq Disaster Recovery/RDF stand (Exhibition Hall) • Product Information (TMF, AutoTMF, RDF and AutoSYNC) • http://nonstop.compaq.com/view.asp?PAGE=RDF_SW • Presenter Contact Information • Harry Scott, Carr Scott Software +1 781.934.0989harry.scott@carrscott.com • Phil Nye, Cardlink Consultants +44 (1565) 614106phil@cardlink.co.uk • Robert Cline, Brown & Co. +1 (617) 878-2708rcline@brownco.com • John Green, BBS AS +47 2289 8070john.green@bbs.no

More Related